Adaptive slices in brain haemorrhage segmentation based on the slic algorithm

Abstract

Traffic accidents have a significant impact on daily life, causing head injuries like skull fractures, brain damage, and so on. Many people fail to follow the safety regulations, such as riding a motorcycle without a helmet. The use of machine learning in brain haemorrhage research is extremely challenging since it involves the collection of patient data from computed tomography (CT) scan images. This study proposes a novel region-based segmentation approach for improving the accuracy and efficiency of CT automated 3D image processing in the analysis of brain injuries. It is quite challenging to create a highly efficient superpixel method which maintains a strategic distance from the segmentation and limited clusters of the pixels in respect to the intensity boundaries. The approach reduces computational costs, and the model achieves 97.79% accuracy in segmenting brain haemorrhage images. This study also guides the direction of future research in this domain.